Enable model.to(device) for int8 weight only quantized model

Summary:
Fix some implementation issue for `int8_wo_quantized_model.to(device)`

Test Plan:
python test/quantization/test_quant_api.py -k test_quantized_model_to_device

Reviewers:

Subscribers:

Tasks:

Tags:

Author: jerryzh168

test/quantization/test_quant_api.py

@@ -620,5 +620,21 @@ def test_quantized_tensor_subclass_save_load(self):
         self.assertEqual(res, ref)
 
 
+    @unittest.skipIf(not TORCH_VERSION_AFTER_2_4, "Test only enabled for 2.4+")
+    @unittest.skipIf(not torch.cuda.is_available(), "Need CUDA available")
+    def test_quantized_model_to_device(self):
+        m = ToyLinearModel().eval().to(torch.bfloat16)
+        m_copy = copy.deepcopy(m)
+        example_inputs = m.example_inputs(dtype=torch.bfloat16, device="cpu")
+
+        quantize_(m, int8_weight_only())
+        ref = m(*example_inputs)
+
+        example_inputs_cuda = (example_inputs[0].to("cuda"),)
+        m.to(device="cuda")
+        cuda_res = m(*example_inputs_cuda)
+        self.assertEqual(cuda_res.cpu(), ref)
+
+
 if __name__ == "__main__":
     unittest.main()

torchao/dtypes/affine_quantized_tensor.py

@@ -259,8 +259,11 @@ def _get_to_kwargs(self, *args, **kwargs):
 
     def to(self, *args, **kwargs):
         kwargs = self._get_to_kwargs(*args, **kwargs)
+        device = kwargs.pop("device")
+        # not supported yet
+        kwargs.pop("memory_format")
         return self.__class__(
-            self.layout_tensor.to(kwargs["device"]),
+            self.layout_tensor.to(device),
             self.block_size,
             self.shape,
             self.quant_min,
@@ -470,8 +473,8 @@ def to(self, *args, **kwargs):
         if device != "cuda" or (isinstance(device, torch.device) and device.type != "cuda"):
             raise ValueError(f"TensorCoreTiledAQTLayout is only available for cuda device")
         return self.__class__(
-            self.packed_weight.to(kwargs["device"]),
-            self.scale_and_zero.to(kwargs["device"]),
+            self.packed_weight.to(device),
+            self.scale_and_zero.to(device),
             self.transposed
         )